Computational Study of Non-planar Hydration Layer and Gas Hydrate Nucleation using Molecular Dynamics and Enhanced Sampling Techniques = 분자동역학 시뮬레이션과 향상된 샘플링 기법을 이용한 비평면 수화층 및 가스 하이드레이트 핵생성에 대한 전산 연구

다국어 초록 (Multilingual Abstract)

Understanding water molecule behavior at non-planar interfaces represents a fundamental research challenge with profound implications for atmospheric aerosol formation, energy storage, and carbon capture technologies. This dissertation systematically investigates the molecular-level structure and dynamics of water at curved interfaces and their critical role in gas hydrate nucleation through advanced molecular dynamics simulations and enhanced sampling techniques.
Characterization of water droplets spanning from nanoscale (128 molecules) to macroscopic systems (71,000 molecules) reveals that interfacial curvature fundamentally modulates local hydration layer properties. Interfacial thickness monotonically increases from 3.0 Å to 4.8 Å with decreasing curvature, while residence times increase from 9.2 ps to 13.0 ps. The fraction of under-coordinated water molecules (forming fewer than three hydrogen bonds) decreases from 53.4% at planar interfaces to 27.8% in highly curved droplets, demonstrating that curvature paradoxically stabilizes a more robust hydrogen bond network. Dynamical analysis reveals that H-bond lifetimes are shortest in small droplets ( ≈ 2.56 ps) and increase toward the planar limit, while OH bond reorientation occurs 10-30% faster at interfaces than in bulk, with free OH groups exhibiting 5-6 times faster dynamics than hydrogen-bonded counterparts.
To address the fundamental rare event problem in gas hydrate nucleation, Forward Flux Sampling (FFS) combined with Mean First Passage Time (MFPT) analysis was implemented to compute CO₂ hydrate nucleation rates and free energy landscapes under experimentally relevant conditions. Using the Mutually Coordinated Guest (MCG) order parameter with adaptive interface placement strategies, this approach successfully overcame the "exceeding the age of the Universe" timescale problem that renders direct molecular dynamics infeasible, enabling simultaneous determination of both kinetic rates and thermodynamic free energy profiles.
This comprehensive investigation establishes quantitative structure property dynamics relationships for curved aqueous interfaces and demonstrates how advanced sampling methodologies can access thermodynamic and kinetic information fundamentally inaccessible to conventional simulation approaches. The molecular-level insights inform development of predictive theories beyond classical nucleation theory and provide essential benchmark data for rational design of nucleation control strategies in carbon sequestration and energy storage applications.

목차 (Table of Contents)

• 1. Introduction 1
• 2. Non-flat Water-Air Interface 8
• 2.1. Molecular Dynamics Simulation Setup 10
• 2.1.1. Simulation Parameters and Solver 10
• 2.1.2. Simulation Configurations for Curvature Study 10
• 2.2. Define Interface 14
• 2.2.1. Gibbs Dividing Surface 14
• 2.2.2. Residence Time of H2O at Interface 20
• 2.3. Structural Properties on Interface vs bulk 24
• 2.3.1. Define Surface Normal Direction 24
• 2.3.2. Dipole Moment Orientation Distribution Analysis 26
• 2.3.3. OH Orientation Distribution Analysis 30
• 2.3.4. Tetrahedral Order Parameters 39
• 2.3.5. Number of H-bond per molecules 42
• 2.4. Dynamical Properties on Interface vs bulk 51
• 2.4.1. Hydrogen Bond Time Correlation Analysis 51
• 2.4.2. Dipole Correlation Analysis 55
• 2.4.3. OH Bond Correlation Analysis 60
• 3. Nucleation of Clathrate Hydrate 68
• 3.1. Collective Variables for Hydrate Nucleation 70
• 3.2. Implementation of Forward Flux Sampling 72
• 3.2.1. Common Force Fields and System Configuration 72
• 3.2.2. Conventional Molecular Dynamics Runs for Preparation Configuration 74
• 3.2.3. Rigorous Definition of States and Selection 75
• 3.2.4. FFS Iteration 76
• 3.2.5. Adaptive Interface Placement Strategy 79
• 3.3. Free Energy Computation via MFPT Analysis 83
• 3.3.1. Theoretical Framework: Fokker-Planck Formalism 83
• 3.3.2. Free Energy Calculation via Trajectory Statistics 85
• 3.3.2.1. Estimation of Steady-State Probability 86
• 3.3.2.2. Estimation of Mean First Passage Time 87
• 3.4. Nucleation Kinetics and Free Energy Landscape 88
• 3.4.1. Determination of Interface Crossing Probabilities 88
• 3.4.2. Mean First Passage Time Analysis 90
• 3.4.3. Steady-State Probability Analysis 93
• 3.4.4. Analysis Free Energy Profile of CO2 Hydrate Nucleation 95
• 4. Conclusion 99